Biological systems approach the limits of miniaturization. Both the proteins and nucleic acids found in biological systems are currently being manipulated using modern biotechnology. Recent progress in protein engineering clearly shows that the principles of the design and production of proteins of almost any desired functionality are now available. These principles have now been applied to the production of completely new proteins designed with detailed knowledge of protein structure and function (1)..sup.1
 FNT .sup.1 See the bibliography list following the specification.
A logical next step in biotechnology is the fabrication of assemblies and devices on the nanometer scale. Since most devices take advantage of the proximity and precise 3D arrangement of individual components, one of the limitations in the fabrication of nanoscale devices has been the inherent lack of specificity in chemical methods for addressing components like bioengineered proteins to precise locations in a 2D array or 3D lattice.
The modular assembly of arrays is easily approached with DNA. Branching through the formation of Watson-Crick paired duplexes in the shape of a Y or an X is now well known (2-5), and Seeman and co-workers have pioneered the assembly of these modules into arrays and lattices with considerable success. Their experiments have demonstrated the feasibility of assembly of 2-D quadrilaterals and 3-D cubes on which more extended structures can be based.
More recently the synthesis of branching dendritimers of single stranded DNA has been reported (6). These components employ controlled introduction of bifunctional phosphoramidite which can cross-link chains in order to assembly precisely defined branched molecules.
While a 2-D or 3-D lattice of B-DNA having considerable complexity can now be constructed (7-8), only a limited number of applications can be envisioned for these structures unless addressable linking of useful components can be achieved. Seeman and co-workers have proposed devices based on the docking of conducting polymers or the attachment of enzymes through an antibody linkage (4,5). To date, antibodies directed against DNA have been difficult to produce and those that have been prepared have only a limited capacity to recognize DNA sequence. Thus, the selectivity of an antibody-based addressing system is questionable. Recognition of an organic hapten might provide more specificity. Here one would introduce a substitution at a prescribed site during synthesis into the DNA. Antibodies to the hapten could be used to decorate the matrix depending on the pattern laid down during synthesis. If the antibodies were engineered to be bifunctional, then they could be used as secondary attachment sites for a second antigen. The disadvantage here is that all hapten moieties are equivalent and thus selective addressing would not be possible unless a series of haptens and antibodies directed to them could be developed. Similar considerations hold for ligand binding systems like avidin-biotin. While a system of distinct haptens and antibodies is possible, developing a set of hapten-phosphoramidites and the series of bifunctional antibodies would be exceedingly time consuming. Moreover, the use of non-covalent linkages in order to achieve addressing would sacrifice stability.
The DNA(cytosine-5)methyltransferases may provide a key advance in addressable linking. The properties of DNA(cytosine-5)methyltransferases that are useful in this context are: (1) the well-characterized DNA sequence specificities of the various bacterial enzymes (9), and (2) the formation of a dihydrocytosine intermediate during catalysis which results in the formation of a covalent complex between a group at the active site and 5-fluorocytosine (5-FdC) at the cytosine methylacceptor in the DNA recognition sequence of the methyltransferase (10-13).
In the biological catalysis carried out by the enzymes (FIG. 1), nucleophilic attack at C-6 of the cytosine ring in the DNA recognition site of the enzyme saturates the 5-6 double bond to produce a dihydrocytosine intermediate activated as a methylacceptor at C-5. The methyltransferase catalyzes methlytransfer to C-5 from S-adenosylmethionine to produce S-adenosylhomocysteine. .beta.-elimination of the enzyme and the hydrogen at C-5 produces 5-methylcytosine and active enzyme. This process cannot be completed when 5FdC is attacked because both the C--C bond to the methyl group and C--F bond at C-5 are too strong to permit abstraction of either the methyl or the fluorine at C-5 as is required for .beta.-elimination to proceed. The net result of enzymatic attack of 5FdC is the production of a stable covalent complex between the enzyme and DNA at specific DNA recognition sites containing 5FdC (16). See FIGS. 1 and 2A. A model of the enzyme covalently bound to an extended DNA molecule is depicted in FIG. 2B.
The recent elucidation of the three-dimensional structure of the HhaI methyltransferase covalently bound to 5FdC at its DNA recognition site shows that the covalent complex is further stabilized by enzyme-DNA contacts which give it its sequence recognition specificity (15).